Tuberculosis (TB) continues to be an important public health problem, with an estimated 9.4 million cases and 1.7 million deaths in 2009. Multidrug resistant TB (MDR-TB, defined as resistance to at least isoniazid and rifampicin), and HIV-associated TB pose important threats to TB control. The high case fatality rate of HIV- associated MDR-TB has major implications for sub-Saharan Africa. The lack of laboratory capacity for Mycobacterium tuberculosis culture and drug susceptibility testing (DST) in resource limited settings poses important challenges to the fight against MDR-TB. In 2008, only 7% of the estimated 440,000 MDR-TB cases worldwide were detected. Xpert MTB/RIF, a novel rapid diagnostic that simultaneously detects M. tuberculosis and rifampicin resistance within two hours, was recently (Dec. 2010) recommended by the WHO as the initial test in those suspected of MDR-TB or HIV-associated TB. The assay detects rifampicin resistance with 95.1% sensitivity and 98.4% specificity, high negative predictive value (99%) but relatively low positive predictive in people with a low pre-test probability of MDR-TB. While rapid diagnosis of rifampicin resistance could revolutionize the fight against MDR-TB, this technological advance in rapid diagnosis will only result in improved patient outcomes if the rapid diagnosis is linked to rapid access to optimal treatment. In 2008, only 11% of MDR-TB cases were appropriately treated. To ensure optimal treatment outcomes and avoid amplification of resistance, knowledge of the complete resistance profile is required. Using Xpert MTB/RIF, clinicians will need to make a treatment decision based on the knowledge of rifampin resistance only, and thus risk the initiation of a suboptimal regimen. To guide the initial standardized management of rifampicin resistance detected by Xpert MTB/RIF, we will comprehensively characterize the phenotypic and genotypic resistance profiles in a large (n=474) number of patients with rifampicin resistant TB. To evaluate the impact of Xpert MTB/RIF on patient outcomes, we will perform implementation research nested within the phased implementation of Xpert MTB/RIF by the South African Department of Health (DOH). We will document treatment decisions made in a cohort of rifampicin resistant TB cases, and compare outcomes (time to culture conversion, amplification of resistance, drug toxicity, and survival rates) during the first 6 months of treatment between 237 patients diagnosed with rifampicin resistance on Xpert MTB/RIF and 237 patients with rifampicin resistance detected on culture based-DST (patients without routine access to Xpert MTB/RIF). The proposed implementation research aims to change the current paradigm for screening, diagnosis and treatment of MDR- TB, by building a robust knowledge base on the resistance profiles of patients diagnosed with rifampicin resistant TB on Xpert MTB/RIF, by quantifying the impact of the assay on amplification of resistance and treatment outcomes in patients with drug resistant TB. The evidence generated will contribute to the successful implementation of this novel assay into routine practice in resource limited, high TB/HIV burden countries. PUBLIC HEALTH RELEVANCE: Multidrug resistant tuberculosis (MDR-TB) is an important public health problem. Xpert MTB/RIF is a novel rapid diagnostic that simultaneously detects M. tuberculosis and rifampicin resistance. We will characterize the phenotypic and genotypic anti- tuberculosis drug resistance profiles in 474 patients with rifampicin resistant TB in South Africa, and quantify the impact of Xpert MTB/RIF on treatment outcomes of patients with drug resistant TB.